Sam Altman states that a crisis in employment due to AI is improbable.

      The chief of OpenAI, while speaking in the Asia-Pacific region, toned down earlier dramatic forecasts of widespread job loss. Current data supports his perspective.

      On Tuesday, OpenAI’s CEO Sam Altman stated that artificial intelligence is unlikely to lead to the extensive employment collapse often referred to in industry jargon as the jobs apocalypse. However, he acknowledged that certain job categories, like customer support, might significantly diminish.

      His comments, reported by Reuters during an Asia-Pacific event, represent a noticeable shift in tone from executives who previously expressed strong concerns regarding AI's impact on labor. Altman has spent much of the past year asserting that customer service positions are "totally, totally gone" in the near future, emphasizing that traditional job skills now have a half-life of two to three years.

      The new perspective, shared during his recent visits to India, Japan, and South Korea, conveys a different message: While there will be substantial turnover in certain sectors, a widespread economic collapse in employment numbers is not expected. This adjustment aligns with the lack of macroeconomic indicators typically associated with a genuine jobs apocalypse.

      The Yale Budget Lab, which has been monitoring the effects of AI on the U.S. job market since ChatGPT's launch, has not observed significant changes in occupational composition or unemployment duration for workers in roles highly exposed to AI, projected until March 2026. Anthropic's usage data included in the lab’s February update did not alter this outlook. Similarly, the Brookings Institution concluded earlier this year that an apocalypse is not imminent.

      Altman has focused more on the "not yet" aspect. During the India AI Impact Summit in February, he informed CNBC-TV18 that some companies are engaging in “AI washing,” attributing layoffs to AI that would have occurred regardless, although actual displacement is beginning to manifest in specific jobs.

      He has been particularly clear about certain sectors: in his opinion, customer service roles conducted via phone or computer will be replaced by AI within the next few years, and those roles will be accomplished more efficiently. Coding practices have already evolved, with engineers now spending less time on code writing and more on architecture, system design, and evaluating AI-generated work.

      This softer macroeconomic framing aligns with some of OpenAI’s policy initiatives. Earlier in 2026, the company released a 13-page policy paper advocating for taxes on automated labor, a national public wealth fund supported by AI companies, and trials of a 32-hour workweek, suggesting considerable labor market disruption is on the horizon.

      Altman's comments in May should be interpreted as a reassessment of the timing and nature of this disruption, rather than a complete denial of it. He envisions less of a sudden break and more of a gradual reshuffling where certain job categories may disappear, others may transform drastically, and the overall employment figures might not change dramatically.

      Altman’s recent travel itinerary reflects his audience's receptivity to this message. He visited Tokyo in the spring to meet with SoftBank CEO Masayoshi Son and Japanese Prime Minister Shigeru Ishiba, and shortly after, he participated in a developer event in Seoul hosted by OpenAI. His talks in the Asia-Pacific have generally emphasized that “new jobs will emerge” more strongly than in his U.S. presentations.

      The next update from the Yale Budget Lab is expected in the coming weeks. Until then, the overall labor statistics and Altman’s interpretation of them remain relatively stable.